Aged five, Annabel Freyberg's daughter was diagnosed with a rare and deadly cancer. Amid the fear, the family drew comfort from finding themselves at the cutting edge of medical research

Information comes at you thick and fast when your child is seriously ill. First are the facts of the disease and the treatments and tests involved. Then come the hospital routines and learning to be your child's nurse at home – administering food and medication via a nose tube, using syringes, giving injections. Much of it is alarming and grim, though the people you meet on the way – doctors, anaesthetists, play specialists, nutritionists, hospital teachers and, above all, nurses – are encouraging; even wonderful.

At the beginning, it's hard to take much in. Few things penetrate the horror of learning that your child has a life-threatening illness. When our usually irrepressible four-year-old daughter, Blossom, fell sick in early 2007, and was eventually diagnosed with advanced cancer, the only positive thing that registered with her father and me was that she was to be transferred to the oncology ward at Great Ormond Street Hospital. If anyone could help her, surely it was GOSH?

Not that we knew much about what went on there, other than its high reputation in treating childhood illness. I imagined it was plushly well-funded. My first impression of the place, however, was of a scruffy, confusingly designed entrance and waiting area. But there was no stinting when it came Blossom's care – even long scans were imaginatively handled to minimise fear or discomfort.

The first afternoon we were given a nightmarish talk about the graveness of Blossom's likely condition, a childhood cancer of the nervous system called neuroblastoma. It's rare – some 100 children get this diagnosis in the UK each year – and deadly in its most aggressive form: 35 die from it annually. A week or so of tests followed to ascertain the extent of the cancer; Blossom had stage four, the worst form. Now treatment could begin, though not necessarily a cure, as the disease often recurs.

It was all shatteringly bleak. Even the treatment to prolong her life was likely to cause damage such as hearing loss and kidney impairment. We pleaded for a shred of good news, but the oncology team rightly refuse to give false hope. In our children's presence I kept a reasonably cheery stance. Out of their sight I was desperate and hysterical.

I agonised over not picking up the symptoms sooner. I hated what was going to be done to her, even though it was her only hope. I sobbed when I saw "normal" families in the street, and avoided everyone I could, but I am endlessly grateful for the kindness and prayers of so many people. My husband was as shaken, but kept his composure, taking his lead from Blossom he said that if she wasn't crushed by this ordeal, neither should we be. And we trusted Great Ormond Street – Blossom was receiving state-of-the-art treatment.

Our consultant, Julia Chisholm, introduced us to clinical trials, explaining that because childhood cancers are far rarer than adult ones, there is fantastic international co-operation and pooling of information. Blossom was eligible to take part in the European International Society of Paediatric Oncology's neuroblastoma study, which has been running for several years. She assured us that the treatment would be exactly the same except for two possible randomisations on the trial: the European or US variety of high-dose chemotherapy and immunotherapy – which wasn't yet available in the UK arm of the trial, as there were delays in processing the new European antibody for use in children. Both elements were some months away. She emphasised that there was no pressure to take part, and that even if we did we were free to back out at any point. I was keen for Blossom to participate. In the middle of so much that was frightening and distressing I found it comforting that her progress would be monitored as something of scientific import – however small her role, even though she wouldn't benefit from it directly.

In the past 30 years, medical advances have resulted in far higher survival rates in children with cancer, a rise of around 10 per cent a decade. Thus, in 1978, five in 10 children were still alive five years after diagnosis, whereas today it is eight in 10. However, the success rate has been patchy. The cure rates are very good for some diseases, such as acute lymphoblastic leukaemia, but there has been little progress in other cancers – for example in some high-grade brain tumours and high-risk neuroblastoma in those over 18 months – hence the need for new approaches.

Encouragingly, Dr Chisholm explained that this trial had already yielded one positive result: the randomised use of a growth factor to stimulate the production of white blood cells during the first course of chemotherapy had proved so successful it was now given to everyone with high-risk neuroblastoma.

Even with international co-operation, large trials take years. This one had started in 2002 and was still open in 2008, after accruing 1,000 patients – 175 a year. Great care is taken with all clinical trials to explain the aims, potential benefits and risks to participants, particularly where children are concerned.

We were given copious literature about each stage of treatment (eight cycles of induction chemotherapy, stem cell collection, surgery, high-dose chemotherapy, stem cell rescue, radiotherapy, possible immunotherapy, cis-retinoic acid) and a timetable. If all went well, there would be six months of intensive treatment followed by six months of oral cis-retinoic acid – designed to mop up any remaining undetectable cancer cells.

I'm sure we were told that it was a phase three trial, but I only registered later that trials came in phases, with one the most experimental (trying out drugs that looked promising in the lab), and three comparing "the best treatment you know against a theoretical suggestion that another might be better", as Anthony Michalski, a consultant paediatric oncologist at GOSH, said.

Patients undergoing treatment for the first time are usually only entered on phase three or four trials (which monitor a drug that has already been licensed). If the cancer returns, the child may be eligible for smaller scale phase one and two trials, testing a new drug's efficacy and side effects.

Most parents – including myself – know little about the different researches into their own child's disease, unless invited on a trial or through media interest in a new report. However, positive results are often put to good use long before a report's publication.

Thus, little did I know that Blossom received her first doses of chemotherapy every ten days rather than every 21 (the standard treatment up to the mid-90s) because of the results of an earlier phase three trial that weren't even published until February 2008. Moreover, the doses of chemotherapy were 1.8 times stronger than previously, as both this and the 10-day interval between them had resulted in a higher survival rate.

Blossom was high-spirited throughout, even when the going was tough, and through constant extra hospitalisations. She enjoyed her relationship with the nurses, and took pride in being stoic when the unpleasant things were done to her. It made the whole thing more bearable as her parents.

For eight weeks from August 2007, Blossom went into GOSH for high-dose chemotherapy and stem cell transplant. It was hard-going for her, but surprisingly jolly, too, with a rash of family birthdays. We often had tea parties on the open-roof playground, with Blossom trailing her feed or medication on a stand. When unattached, she donned wings and did a birthday dance.

In the first few days back we were told of another trial (run by the bone-marrow transplant team) and agreed she should take part. It studied the effectiveness of an anti-clotting agent in preventing and treating a life-threatening liver disease that can occur after high-dose therapy. The computer decreed that Blossom would be in the group that didn't get the drug. This was disappointing – though mercifully she didn't have liver problems. I later heard of another parent's distress at her child having to take yet another drug for weeks without any apparent benefit. That is the downside of a clinical trial, and one of the reasons some parents opt out.

Trials are not core NHS work. Doing them requires, in Dr Michalski's words, a "different mindset to usual on the ward", and dedicated staff who are "good clinical nurses and good data managers; data monitoring requirements are strict, and without clinical research nurses, studies don't happen".

Karen Howe became GOSH's first oncology research nurse in 2000. She has opened dozens of phase one and two trials and other supportive care and pharmacology studies. Tina Say stood in for Karen in 2007, and has recently become one of two leukaemia clinical nurse specialists, working on phase three trials. Her post and that of two new clinical research nurses for phase one and two trials will be funded for three years by the 2009 Kiss It Better appeal, which ran through last month.

Kiss It Better was founded in 2004 by Carmel Allen, a former journalist whose daughter was treated for neuroblastoma at GOSH, and who wanted to turn her experience into something positive. Along with University College London Hospital (which takes children aged 13 and over; GOSH takes newborns to 12 year olds), GOSH is part of the largest centre for children with cancer and leukaemia in Europe, and the third largest in the world.

New drugs for childhood cancers depend on the passion of doctors and scientists – there's little financial benefit because only one in 600 children is affected by cancer compared to one in three adults. Even so, there are "more and more trials coming through," Dr Michalski states. And most children with cancer are eligible for trials.

Blossom's cancer is currently in remission and she has been back at school for more than a year. She's mad about swimming, fairies and science club, and longs for a puppy. When she visited GOSH last month, I was delighted to see that the waiting area has been smartly revamped with a wall map showing each ward. Appearances may not matter as much as science, but they're jolly useful all the same.